In: Statistics and Probability
SUMMARY OUTPUT |
|||||||||
Regression Statistics |
|||||||||
Multiple R |
0.396235 |
||||||||
R Square |
0.157002 |
||||||||
Adjusted R Square |
0.156262 |
||||||||
Standard Error |
18.42647 |
||||||||
Observations |
1142 |
||||||||
ANOVA |
|||||||||
df |
SS |
MS |
F |
Significance F |
|||||
Regression |
1 |
72088.71 |
72088.71 |
212.3161 |
3.12E-44 |
||||
Residual |
1140 |
387069.6 |
339.5348 |
||||||
Total |
1141 |
459158.4 |
|||||||
Coefficients |
Standard Error |
t Stat |
P-value |
Lower 95% |
Upper 95% |
Lower 95.0% |
Upper 95.0% |
||
Intercept |
26.35917 |
0.803163 |
32.8192 |
7.4E-167 |
24.78333 |
27.93501 |
24.78333 |
27.93501 |
|
X Variable 1 |
0.000213 |
1.46E-05 |
14.57107 |
3.12E-44 |
0.000184 |
0.000242 |
0.000184 |
0.000242 |
|
a. Write the reqression equation.
Solution:
Given:
We are given regression analysis output of variables Income and Number of work hours per week.
Part a) Write the regression equation.
General regression equation is given by:
y = b0 + b1 * x
Here y = dependent variable = Income , x = independent variable = Number of work hours per week
b0 = Intercept = 26.35917
b1 = Slope = 0.000213
Thus regression equation is:
Income = 26.35917 + 0.000213 * Number of work hours per week
Part b) Discuss the statistical significance of the model using the appropriate regression statistic at a 95% level of confidence , that is for 1 - 0.95 = 0.05 level of significance.
ANOVA table is used to test the significance of overall model.
From ANOVA we can see: Significance F = 3.12E-44
3.12E-44 is scientific number and E-44 means we have to move 44 decimal places to left of 3.12
Thus rounding this number to 4 decimal places we get:
Significance F = 0.0000
That is: P-value = 0.0000
Since P-value = 0.0000 < 0.05 level of significance , we conclude that the fitted model is significant.
Part c) Discuss the statistical significance of the coefficient for the independent variable using the appropriate regression statistic at a 95% level of confidence.
From P-value column, we can see P-value for slope coefficient is 3.12E-44 = 0.0000
Since P-value = 0.0000 < 0.05 level of significance , the coefficient for the independent variable is statistically significant.
Part d) Interpret the coefficient for the independent variable.
Intercept is the value of dependent variable y , when x = 0.
Here y is income and x is Number of work hours per week.
If Number of work hours per week is 0, then there would be no possibility of Income.
Thus intercept is meaningless in this context.
That is: if Number of work hours per week is 0 , then it is meaningless to have income of 26.35917 .
Slope: Slope is the coefficient of independent variable. It gives amount of change in dependent variable when independent variable changed by one unit.
Thus we have Slope = 0.000213.
Thus if we increase number of hour per week by one hour, then there is increase of Income by the amount 0.000213.
Part e) What percentage of the observed variation in income is explained by the model?
R-square gives the amount variation in dependent variable which is explained by model ( by independent variable)
Here R-square = 0.157002 = 15.7%
Thus About 15.7% of the observed variation in income is explained by the model.
Part f) Predict the value of a person’s income who works 50 hours a week, using this regression model.
Put Number of work hours per week = 50 in regression equation obtained in part a)
Income = 26.35917 + 0.000213 * Number of work hours per week
Income = 26.35917 + 0.000213 * 50
Income = 26.35917 + 0.01065
Income = 26.36982
Thus the predicted value of a person’s income who works 50 hours a week is 26.36982